Uploaded image for project: 'Teiid'
  1. Teiid
  2. TEIID-4727

Aggregation is performed incorrectly for JOIN queries

XMLWordPrintable

    • Hide

      1. In postgresql create the tbl_1 table by the following script:

      CREATE TABLE tbl_1
(
  a integer,
  b bigdecimal
);
 
INSERT INTO tbl_1(a, b) VALUES (1, 1);
INSERT INTO tbl_1(a, b) VALUES (1, 2);
INSERT INTO tbl_1(a, b) VALUES (1, 2);
INSERT INTO tbl_1(a, b) VALUES (1, 3);

      2. Add postgresql database configuration in standalone-teiid.xml

      <datasource jndi-name="java:/test_pg" pool-name="test_pg" enabled="true" use-java-context="true">
                    <connection-url>jdbc:postgresql://localhost:5432/test?charSet=utf8</connection-url>
                    <driver-class>org.postgresql.Driver</driver-class>
                    <driver>org.postgresql</driver>
                    <pool>
                        <min-pool-size>2</min-pool-size>
                        <max-pool-size>70</max-pool-size>
                        <prefill>false</prefill>
                        <use-strict-min>false</use-strict-min>
                        <flush-strategy>FailingConnectionOnly</flush-strategy>
                    </pool>
                    <security>
                        <user-name>postgres</user-name>
                        <password>xxxxxx</password>
                    </security>
                    <validation>
                        <check-valid-connection-sql>select 0</check-valid-connection-sql>
                    </validation>
                    <timeout>
                        <blocking-timeout-millis>120000</blocking-timeout-millis>
                        <idle-timeout-minutes>5</idle-timeout-minutes>
                    </timeout>
                </datasource>

      3. Add in test-vdb.xml java:/test_pg configured in previous step as datasource:

      <model name="test_pg">
        <property name="importer.useFullSchemaName" value="false"/>
	<property name="importer.tableTypes" value="TABLE,VIEW"/>
	<property name="importer.importKeys" value="false"/>
        <source name="test_pg" translator-name="myPg" connection-jndi-name="java:/test_pg"/>
    </model>

      4. Configure in test-vdb.xml the following virtual views:

      <model visible = "true" type = "VIRTUAL" name = "views">
        <metadata type = "DDL"><![CDATA[
          CREATE virtual view v1 as select * from test_pg.tbl_1;
          CREATE virtual view v2 as select 1 as a, 1 as b
        ]]>
        </metadata>
</model>

      5. Configure in test-vdb.xml the following virtual procedure:

      <model visible = "true" type = "VIRTUAL" name = "procs">
        <metadata type = "DDL"><![CDATA[
          create procedure pr() returns (a integer, b integer) as 
          begin
	     select 1, 1;
          end        
        ]]>
        </metadata>
</model>

      6. Run test queries mentioned in description field:

      --Run test queries and check result, which should be:
--2
--1
--1
-- result is correct
select count(v2.b) from views.v1 join views.v2 on true group by v1.b;

-- result is correct
select count(v2.b) from views.v1 left join views.v2 on true group by v1.b;

-- result is incorrect
select count(v2.b) from views.v1 right join views.v2 on true group by v1.b;

-- result is incorrect
select count(v2.b) from views.v1 join (call procs.pr()) v2 on true group by v1.b;

-- result is incorrect
select count(v2.b) from views.v1 left join (call procs.pr()) v2 on true group by v1.b;

-- result is incorrect
select count(v2.b) from views.v1 right join (call procs.pr()) v2 on true group by v1.b;
      Show
      1. In postgresql create the tbl_1 table by the following script: CREATE TABLE tbl_1 ( a integer , b bigdecimal ); INSERT INTO tbl_1( a , b) VALUES (1, 1); INSERT INTO tbl_1( a , b) VALUES (1, 2); INSERT INTO tbl_1( a , b) VALUES (1, 2); INSERT INTO tbl_1( a , b) VALUES (1, 3); 2. Add postgresql database configuration in standalone-teiid.xml <datasource jndi-name= "java:/test_pg" pool-name= "test_pg" enabled= "true" use-java-context= "true" > <connection-url> jdbc:postgresql://localhost:5432/test?charSet=utf8 </connection-url> <driver-class> org.postgresql.Driver </driver-class> <driver> org.postgresql </driver> <pool> <min-pool-size> 2 </min-pool-size> <max-pool-size> 70 </max-pool-size> <prefill> false </prefill> <use-strict-min> false </use-strict-min> <flush-strategy> FailingConnectionOnly </flush-strategy> </pool> <security> <user-name> postgres </user-name> <password> xxxxxx </password> </security> <validation> <check-valid-connection-sql> select 0 </check-valid-connection-sql> </validation> <timeout> <blocking-timeout-millis> 120000 </blocking-timeout-millis> <idle-timeout-minutes> 5 </idle-timeout-minutes> </timeout> </datasource> 3. Add in test-vdb.xml java:/test_pg configured in previous step as datasource: <model name= "test_pg" > <property name= "importer.useFullSchemaName" value= "false" /> <property name= "importer.tableTypes" value= "TABLE,VIEW" /> <property name= "importer.importKeys" value= "false" /> <source name= "test_pg" translator-name= "myPg" connection-jndi-name= "java:/test_pg" /> </model> 4. Configure in test-vdb.xml the following virtual views: <model visible = "true" type = "VIRTUAL" name = "views" > <metadata type = "DDL" > <![CDATA[ CREATE virtual view v1 as select * from test_pg.tbl_1; CREATE virtual view v2 as select 1 as a, 1 as b ]]> </metadata> </model> 5. Configure in test-vdb.xml the following virtual procedure: <model visible = "true" type = "VIRTUAL" name = "procs" > <metadata type = "DDL" > <![CDATA[ create procedure pr() returns (a integer, b integer) as begin select 1, 1; end ]]> </metadata> </model> 6. Run test queries mentioned in description field: --Run test queries and check result , which should be: --2 --1 --1 -- result is correct select count (v2.b) from views.v1 join views.v2 on true group by v1.b; -- result is correct select count (v2.b) from views.v1 left join views.v2 on true group by v1.b; -- result is incorrect select count (v2.b) from views.v1 right join views.v2 on true group by v1.b; -- result is incorrect select count (v2.b) from views.v1 join ( call procs.pr()) v2 on true group by v1.b; -- result is incorrect select count (v2.b) from views.v1 left join ( call procs.pr()) v2 on true group by v1.b; -- result is incorrect select count (v2.b) from views.v1 right join ( call procs.pr()) v2 on true group by v1.b;

      Aggregation is performed incorrectly for JOIN queries, like this

      select count(v2.b) from views.v1 right join views.v2 on true group by v1.b;

select count(v2.b) from views.v1 join (call procs.pr()) v2 on true group by v1.b;

select count(v2.b) from views.v1 left join (call procs.pr()) v2 on true group by v1.b;

select count(v2.b) from views.v1 right join (call procs.pr()) v2 on true group by v1.b;

      showing as a result:

      1
1
1

      though these ones work correctly:

      select count(v2.b) from views.v1 join views.v2 on true group by v1.b;

select count(v2.b) from views.v1 left join views.v2 on true group by v1.b;

      showing

      1
1
2

              rhn-engineering-shawkins Steven Hawkins
              dalex005 Dmitrii Pogorelov
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Created:
                Updated:
                Resolved: